VFD - Frequent Frequency changes

I've got a question, maybe an issue, but more just curiosity.

I have an air compressor, 100HP, that runs with VFD contorls. It is the only one in the loop so I don't have any worries about a second VFD driven air compressor causing issues.

When the factory is in low air usage, the VFD is at a stable output frequency. When the air demand goes high, the VFD is also stable. But when the air usage is somewhere in the middle, the VFD tends to make frequent frequency changes. It changes rapidly from 30 to 60 Hz output. I'm thinking that some parameter in the controller's PID is not tuned right, but don't know enough about it to attempt to tune it myself. The service technicians did not think it was a concern.

My curiosity is whether or not these rapid frequency changes will wear out the components any faster? We recently had the bearings fail in the motor and had to repair them. I don't have any CFM numbers or anything like that, and no way currently to measure them.

The real concern is if the pressure stays within the allowable tolerance for your process. If it's staying constant, then obviously there is a large flow demand somewhere in the system.

Rapid changes only really matter if they are stressing something. Are you hitting > 100% FLA on the motor when accelerating?

System pressure is pretty constant. So the factory's air system doesn't actually see any issues. I was just looking at it from the point of view that the VFD should be able to find a happy medium in the CFM output of that compressor and stick to it rather than rapidly climbing, only to immediately dial itself back down, over and over.

The amps on the motor do not appear to be going over FLA. I have only viewed snippets of the whole time, but the times I have viewed, it wasn't overloaded.

I just realized that I forgot a piece in the first message. These 30-60 Hz changes repeat every second. Constantly for 15 minutes at a time. Before usage changes in the plant and the compressor dials back to a lower or stablizes at a higher output.

Are you able to get online and view the logic for that PID? Is that compressor driven by a PLC or a PID controller?

You "should" be able to play with the PID as long as you KNOW where the original number was of P, I, and D. That way you can put it back to that number if something is out of wack.

Of course don't mess with it if you feel like you could shut something down!

I too am becoming more curious about PID controllers as well.

It's some sort of proprietary controller from the air compressor manufacturer with a Siemens VFD. I'm not able to actually get in and view anything in their controller logic, just the parameters they make available through the interface.

The only reason I'm really thinking about it is because my office shares a wall with the compressor and that's all I hear all day long. The "Wrrrrr" noise of the compressor ramping up and down constantly.

Ahh got ya. It seems like the signal to the drive is bouncing around and the drive is reacting to that signal (as you already know). But you'd think it would have better debounce coding than that.

If you ever get an answer, and can remember to reply; I'd like to know what it is...

It's not noise or a debounce issue. I'm pretty sure it's just the PID is tuned tightly. The amount the output is changing with small changes in PV is huge. From what little I understand about PIDs, I would think it's the P or D portions that would cause it.

But my real concern is whether or not the frequent speed changes put any more wear on the motor or compressor.

The fact that the pressure stays constant indicates that the system is working well and just dealing with rapid load changes. It's not surprising that it is stable at low load. It's also not surprising that it's stable at high load. In both cases, the load isn't changing. If you have a lot of loads that are starting up and shutting down, the compressor is going to work hard to hold the pressure, that means step changes in speed to deal with step changes in flow.

Now if the pressure is oscillating at the same time, you may have a tuning issue.

I guess it's a non issue, just didn't know much about it.

Prior, when I first noticed it, the maintenance crew had hooked up the compressor without a tank. The only volume of air was what existed in the pipes. When the system was connected like that, the VFD was constantly fluctuating and the pressure was as well, as soon as the VFD ramped up, it went over pressure, and as soon as it slowed down it went under pressure. Didn't matter what the demand was.

The tanks stabilized the pressure. I just assumed it would also stabilize the VFD command. But I think I just didn't take into account that my demand really is changing that often.

I think I'm going to invest in some metering equipment and start recording what my CFM usage is over time. That will likely tell me a lot more than just looking at the VFD output of my compressor.

What about using a volume tank? We are in the midst of installing a compressor bank into our facility. 3 250HP compressors, with one driven by a VFD. Supposedly the VFD driven compressor will be a master compressor that will control off demand, and if unable to maintain demand, a second FVNR compressor will kick in, and then the master will slow. They are supposed to feed into a very large valume tank. We then have two airlines from the volume tank, that have Ingersoll-Rand Pressure Controllers, that have their own built in PID control. Kind of curious about your system, because I have questions when our system is commissioned.

There are two, 400 gallon tanks in the system. One at the compressor and the second tank near the end of the line at the other end of the factory. Introducing those into the system is what stabilized the pressure after the initial installation.

I would consider it an issue. The rapid changes will cause excessive cylinder wall wear and shorten bearing life. Also with the constant motor acceleration, motor windings may be heating up.

One of the possible causes could be poor sensor location. If the pressure sensor is located on the compressor discharge pipe or the system outlet pipe to the field, review locating it to the pressure vessel.

I would consider it a problem also. You should see pretty slow changes with those tanks in there. If you can't get into the PID, you can try to change the accel and decel values in the drive.

The tanks addition essentially added Integral into the PID...

So, either you can increase the Integral via mechanical or software to resolve the issue.

Again, as long as the ramp rate does not push the motor above FLA, it should not cause any undue damage. Rapid load changes will have the same effect.

The oscillations are not good for the motor or the compressor. And considerable energy is being wasted in the process too. If you absolutely cannot gain access to the loop tuning but have access to the line that the air pressure sensor is on, I'd consider adding an orifice in the sensing line and maybe even a small accumulator tank to slow the system response. Less is more if you do this. If you overly deaden the response the system air pressure will not be held stable.

That small accumulator tank would go between the orifice and the pressure sensor.